1 /* gnu/regexp/RESyntax.java
2 Copyright (C) 2006 Free Software Foundation, Inc.
4 This file is part of GNU Classpath.
6 GNU Classpath is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 2, or (at your option)
11 GNU Classpath is distributed in the hope that it will be useful, but
12 WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 General Public License for more details.
16 You should have received a copy of the GNU General Public License
17 along with GNU Classpath; see the file COPYING. If not, write to the
18 Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
21 Linking this library statically or dynamically with other modules is
22 making a combined work based on this library. Thus, the terms and
23 conditions of the GNU General Public License cover the whole
26 As a special exception, the copyright holders of this library give you
27 permission to link this library with independent modules to produce an
28 executable, regardless of the license terms of these independent
29 modules, and to copy and distribute the resulting executable under
30 terms of your choice, provided that you also meet, for each linked
31 independent module, the terms and conditions of the license of that
32 module. An independent module is a module which is not derived from
33 or based on this library. If you modify this library, you may extend
34 this exception to your version of the library, but you are not
35 obligated to do so. If you do not wish to do so, delete this
36 exception statement from your version. */
40 import java
.io
.Serializable
;
41 import java
.util
.BitSet
;
44 * An RESyntax specifies the way a regular expression will be compiled.
45 * This class provides a number of predefined useful constants for
46 * emulating popular regular expression syntaxes. Additionally the
47 * user may construct his or her own syntax, using any combination of the
48 * syntax bit constants. The syntax is an optional argument to any of the
49 * matching methods on class RE.
51 * @author <A HREF="mailto:wes@cacas.org">Wes Biggs</A>
54 public final class RESyntax
implements Serializable
{
55 static final String DEFAULT_LINE_SEPARATOR
= System
.getProperty("line.separator");
57 private static final String SYNTAX_IS_FINAL
= RE
.getLocalizedMessage("syntax.final");
61 // true for the constant defined syntaxes
62 private boolean isFinal
= false;
64 private String lineSeparator
= DEFAULT_LINE_SEPARATOR
;
66 // Values for constants are bit indexes
69 * Syntax bit. Backslash is an escape character in lists.
71 public static final int RE_BACKSLASH_ESCAPE_IN_LISTS
= 0;
74 * Syntax bit. Use \? instead of ? and \+ instead of +.
76 public static final int RE_BK_PLUS_QM
= 1;
79 * Syntax bit. POSIX character classes ([:...:]) in lists are allowed.
81 public static final int RE_CHAR_CLASSES
= 2;
84 * Syntax bit. ^ and $ are special everywhere.
85 * <B>Not implemented.</B>
87 public static final int RE_CONTEXT_INDEP_ANCHORS
= 3;
90 * Syntax bit. Repetition operators are only special in valid positions.
91 * <B>Not implemented.</B>
93 public static final int RE_CONTEXT_INDEP_OPS
= 4;
96 * Syntax bit. Repetition and alternation operators are invalid
97 * at start and end of pattern and other places.
98 * <B>Not implemented</B>.
100 public static final int RE_CONTEXT_INVALID_OPS
= 5;
103 * Syntax bit. Match-any-character operator (.) matches a newline.
105 public static final int RE_DOT_NEWLINE
= 6;
108 * Syntax bit. Match-any-character operator (.) does not match a null.
110 public static final int RE_DOT_NOT_NULL
= 7;
113 * Syntax bit. Intervals ({x}, {x,}, {x,y}) are allowed.
115 public static final int RE_INTERVALS
= 8;
118 * Syntax bit. No alternation (|), match one-or-more (+), or
119 * match zero-or-one (?) operators.
121 public static final int RE_LIMITED_OPS
= 9;
124 * Syntax bit. Newline is an alternation operator.
126 public static final int RE_NEWLINE_ALT
= 10; // impl.
129 * Syntax bit. Intervals use { } instead of \{ \}
131 public static final int RE_NO_BK_BRACES
= 11;
134 * Syntax bit. Grouping uses ( ) instead of \( \).
136 public static final int RE_NO_BK_PARENS
= 12;
139 * Syntax bit. Backreferences not allowed.
141 public static final int RE_NO_BK_REFS
= 13;
144 * Syntax bit. Alternation uses | instead of \|
146 public static final int RE_NO_BK_VBAR
= 14;
149 * Syntax bit. <B>Not implemented</B>.
151 public static final int RE_NO_EMPTY_RANGES
= 15;
154 * Syntax bit. An unmatched right parenthesis (')' or '\)', depending
155 * on RE_NO_BK_PARENS) will throw an exception when compiling.
157 public static final int RE_UNMATCHED_RIGHT_PAREN_ORD
= 16;
160 * Syntax bit. <B>Not implemented.</B>
162 public static final int RE_HAT_LISTS_NOT_NEWLINE
= 17;
165 * Syntax bit. Stingy matching is allowed (+?, *?, ??, {x,y}?).
167 public static final int RE_STINGY_OPS
= 18;
170 * Syntax bit. Allow character class escapes (\d, \D, \s, \S, \w, \W).
172 public static final int RE_CHAR_CLASS_ESCAPES
= 19;
175 * Syntax bit. Allow use of (?:xxx) grouping (subexpression is not saved).
177 public static final int RE_PURE_GROUPING
= 20;
180 * Syntax bit. Allow use of (?=xxx) and (?!xxx) apply the subexpression
181 * to the text following the current position without consuming that text.
183 public static final int RE_LOOKAHEAD
= 21;
186 * Syntax bit. Allow beginning- and end-of-string anchors (\A, \Z).
188 public static final int RE_STRING_ANCHORS
= 22;
191 * Syntax bit. Allow embedded comments, (?#comment), as in Perl5.
193 public static final int RE_COMMENTS
= 23;
196 * Syntax bit. Allow character class escapes within lists, as in Perl5.
198 public static final int RE_CHAR_CLASS_ESC_IN_LISTS
= 24;
201 * Syntax bit. Possessive matching is allowed (++, *+, ?+, {x,y}+).
203 public static final int RE_POSSESSIVE_OPS
= 25;
206 * Syntax bit. Allow embedded flags, (?is-x), as in Perl5.
208 public static final int RE_EMBEDDED_FLAGS
= 26;
211 * Syntax bit. Allow octal char (\0377), as in Perl5.
213 public static final int RE_OCTAL_CHAR
= 27;
216 * Syntax bit. Allow hex char (\x1b), as in Perl5.
218 public static final int RE_HEX_CHAR
= 28;
221 * Syntax bit. Allow Unicode char (\u1234), as in Java 1.4.
223 public static final int RE_UNICODE_CHAR
= 29;
226 * Syntax bit. Allow named property (\p{P}, \P{p}), as in Perl5.
228 public static final int RE_NAMED_PROPERTY
= 30;
231 * Syntax bit. Allow nested characterclass ([a-z&&[^p-r]]), as in Java 1.4.
233 public static final int RE_NESTED_CHARCLASS
= 31;
235 private static final int BIT_TOTAL
= 32;
239 * Emulates regular expression support in the awk utility.
241 public static final RESyntax RE_SYNTAX_AWK
;
245 * Emulates regular expression support in the ed utility.
247 public static final RESyntax RE_SYNTAX_ED
;
251 * Emulates regular expression support in the egrep utility.
253 public static final RESyntax RE_SYNTAX_EGREP
;
257 * Emulates regular expression support in the GNU Emacs editor.
259 public static final RESyntax RE_SYNTAX_EMACS
;
263 * Emulates regular expression support in the grep utility.
265 public static final RESyntax RE_SYNTAX_GREP
;
269 * Emulates regular expression support in the POSIX awk specification.
271 public static final RESyntax RE_SYNTAX_POSIX_AWK
;
275 * Emulates POSIX basic regular expression support.
277 public static final RESyntax RE_SYNTAX_POSIX_BASIC
;
281 * Emulates regular expression support in the POSIX egrep specification.
283 public static final RESyntax RE_SYNTAX_POSIX_EGREP
;
287 * Emulates POSIX extended regular expression support.
289 public static final RESyntax RE_SYNTAX_POSIX_EXTENDED
;
293 * Emulates POSIX basic minimal regular expressions.
295 public static final RESyntax RE_SYNTAX_POSIX_MINIMAL_BASIC
;
299 * Emulates POSIX extended minimal regular expressions.
301 public static final RESyntax RE_SYNTAX_POSIX_MINIMAL_EXTENDED
;
305 * Emulates regular expression support in the sed utility.
307 public static final RESyntax RE_SYNTAX_SED
;
311 * Emulates regular expression support in Larry Wall's perl, version 4,
313 public static final RESyntax RE_SYNTAX_PERL4
;
317 * Emulates regular expression support in Larry Wall's perl, version 4,
318 * using single line mode (/s modifier).
320 public static final RESyntax RE_SYNTAX_PERL4_S
; // single line mode (/s)
324 * Emulates regular expression support in Larry Wall's perl, version 5.
326 public static final RESyntax RE_SYNTAX_PERL5
;
330 * Emulates regular expression support in Larry Wall's perl, version 5,
331 * using single line mode (/s modifier).
333 public static final RESyntax RE_SYNTAX_PERL5_S
;
337 * Emulates regular expression support in Java 1.4's java.util.regex
340 public static final RESyntax RE_SYNTAX_JAVA_1_4
;
345 RE_SYNTAX_EMACS
= new RESyntax().makeFinal();
347 RESyntax RE_SYNTAX_POSIX_COMMON
= new RESyntax()
348 .set(RE_CHAR_CLASSES
)
350 .set(RE_DOT_NOT_NULL
)
352 .set(RE_NO_EMPTY_RANGES
)
355 RE_SYNTAX_POSIX_BASIC
= new RESyntax(RE_SYNTAX_POSIX_COMMON
)
359 RE_SYNTAX_POSIX_EXTENDED
= new RESyntax(RE_SYNTAX_POSIX_COMMON
)
360 .set(RE_CONTEXT_INDEP_ANCHORS
)
361 .set(RE_CONTEXT_INDEP_OPS
)
362 .set(RE_NO_BK_BRACES
)
363 .set(RE_NO_BK_PARENS
)
365 .set(RE_UNMATCHED_RIGHT_PAREN_ORD
)
368 RE_SYNTAX_AWK
= new RESyntax()
369 .set(RE_BACKSLASH_ESCAPE_IN_LISTS
)
370 .set(RE_DOT_NOT_NULL
)
371 .set(RE_NO_BK_PARENS
)
374 .set(RE_NO_EMPTY_RANGES
)
375 .set(RE_UNMATCHED_RIGHT_PAREN_ORD
)
378 RE_SYNTAX_POSIX_AWK
= new RESyntax(RE_SYNTAX_POSIX_EXTENDED
)
379 .set(RE_BACKSLASH_ESCAPE_IN_LISTS
)
382 RE_SYNTAX_GREP
= new RESyntax()
384 .set(RE_CHAR_CLASSES
)
385 .set(RE_HAT_LISTS_NOT_NEWLINE
)
390 RE_SYNTAX_EGREP
= new RESyntax()
391 .set(RE_CHAR_CLASSES
)
392 .set(RE_CONTEXT_INDEP_ANCHORS
)
393 .set(RE_CONTEXT_INDEP_OPS
)
394 .set(RE_HAT_LISTS_NOT_NEWLINE
)
396 .set(RE_NO_BK_PARENS
)
400 RE_SYNTAX_POSIX_EGREP
= new RESyntax(RE_SYNTAX_EGREP
)
402 .set(RE_NO_BK_BRACES
)
405 /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */
407 RE_SYNTAX_ED
= new RESyntax(RE_SYNTAX_POSIX_BASIC
)
410 RE_SYNTAX_SED
= new RESyntax(RE_SYNTAX_POSIX_BASIC
)
413 RE_SYNTAX_POSIX_MINIMAL_BASIC
= new RESyntax(RE_SYNTAX_POSIX_COMMON
)
417 /* Differs from RE_SYNTAX_POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS
418 replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */
420 RE_SYNTAX_POSIX_MINIMAL_EXTENDED
= new RESyntax(RE_SYNTAX_POSIX_COMMON
)
421 .set(RE_CONTEXT_INDEP_ANCHORS
)
422 .set(RE_CONTEXT_INVALID_OPS
)
423 .set(RE_NO_BK_BRACES
)
424 .set(RE_NO_BK_PARENS
)
427 .set(RE_UNMATCHED_RIGHT_PAREN_ORD
)
430 /* There is no official Perl spec, but here's a "best guess" */
432 RE_SYNTAX_PERL4
= new RESyntax()
433 .set(RE_BACKSLASH_ESCAPE_IN_LISTS
)
434 .set(RE_CONTEXT_INDEP_ANCHORS
)
435 .set(RE_CONTEXT_INDEP_OPS
) // except for '{', apparently
437 .set(RE_NO_BK_BRACES
)
438 .set(RE_NO_BK_PARENS
)
440 .set(RE_NO_EMPTY_RANGES
)
441 .set(RE_CHAR_CLASS_ESCAPES
) // \d,\D,\w,\W,\s,\S
444 RE_SYNTAX_PERL4_S
= new RESyntax(RE_SYNTAX_PERL4
)
448 RE_SYNTAX_PERL5
= new RESyntax(RE_SYNTAX_PERL4
)
449 .set(RE_PURE_GROUPING
) // (?:)
450 .set(RE_STINGY_OPS
) // *?,??,+?,{}?
451 .set(RE_LOOKAHEAD
) // (?=)(?!)
452 .set(RE_STRING_ANCHORS
) // \A,\Z
453 .set(RE_CHAR_CLASS_ESC_IN_LISTS
)// \d,\D,\w,\W,\s,\S within []
454 .set(RE_COMMENTS
) // (?#)
455 .set(RE_EMBEDDED_FLAGS
) // (?imsx-imsx)
456 .set(RE_OCTAL_CHAR
) // \0377
457 .set(RE_HEX_CHAR
) // \x1b
458 .set(RE_NAMED_PROPERTY
) // \p{prop}, \P{prop}
461 RE_SYNTAX_PERL5_S
= new RESyntax(RE_SYNTAX_PERL5
)
465 RE_SYNTAX_JAVA_1_4
= new RESyntax(RE_SYNTAX_PERL5
)
467 .set(RE_POSSESSIVE_OPS
) // *+,?+,++,{}+
468 .set(RE_UNICODE_CHAR
) // \u1234
469 .set(RE_NESTED_CHARCLASS
) // [a-z&&[^p-r]]
474 * Construct a new syntax object with all bits turned off.
475 * This is equivalent to RE_SYNTAX_EMACS.
478 bits
= new BitSet(BIT_TOTAL
);
482 * Called internally when constructing predefined syntaxes
483 * so their interpretation cannot vary. Conceivably useful
484 * for your syntaxes as well. Causes IllegalAccessError to
485 * be thrown if any attempt to modify the syntax is made.
487 * @return this object for convenient chaining
489 public RESyntax
makeFinal() {
495 * Construct a new syntax object with all bits set the same
496 * as the other syntax.
498 public RESyntax(RESyntax other
) {
499 bits
= (BitSet
) other
.bits
.clone();
503 * Check if a given bit is set in this syntax.
505 public boolean get(int index
) {
506 return bits
.get(index
);
510 * Set a given bit in this syntax.
512 * @param index the constant (RESyntax.RE_xxx) bit to set.
513 * @return a reference to this object for easy chaining.
515 public RESyntax
set(int index
) {
516 if (isFinal
) throw new IllegalAccessError(SYNTAX_IS_FINAL
);
522 * Clear a given bit in this syntax.
524 * @param index the constant (RESyntax.RE_xxx) bit to clear.
525 * @return a reference to this object for easy chaining.
527 public RESyntax
clear(int index
) {
528 if (isFinal
) throw new IllegalAccessError(SYNTAX_IS_FINAL
);
534 * Changes the line separator string for regular expressions
535 * created using this RESyntax. The default separator is the
536 * value returned by the system property "line.separator", which
537 * should be correct when reading platform-specific files from a
538 * filesystem. However, many programs may collect input from
539 * sources where the line separator is differently specified (for
540 * example, in the applet environment, the text box widget
541 * interprets line breaks as single-character newlines,
542 * regardless of the host platform.
544 * Note that setting the line separator to a character or
545 * characters that have specific meaning within the current syntax
546 * can cause unexpected chronosynclastic infundibula.
548 * @return this object for convenient chaining
550 public RESyntax
setLineSeparator(String aSeparator
) {
551 if (isFinal
) throw new IllegalAccessError(SYNTAX_IS_FINAL
);
552 lineSeparator
= aSeparator
;
557 * Returns the currently active line separator string. The default
558 * is the platform-dependent system property "line.separator".
560 public String
getLineSeparator() {
561 return lineSeparator
;